Research On Spatio-Temporal Data Management And Visualization Of Urban Traffic

With the gradual improvement of traffic network construction,the wide application of intelligent sensing technology and communication technology in traffic system,the scale of available traffic data keeps expanding and the temporal and spatial complexity gradually increases,which puts forward new requirements for data management and application.Visualization is an important means of mining the hidden value of data and analyzing the characteristics and distribution rules of data by transforming data into graphs,charts or images.In addition to spatial characteristics,traffic spatio-temporal data is different from other spatio-temporal data,such as timeliness,subjectivity,fluidity and reproducibility.Common visualization methods have certain limitations in some traffic analysis scenarios.The visualization technology of urban traffic spatio-temporal data is studied,and the specific content is as followsBased on the analysis of static urban traffic network facilities data and dynamic traffic flow data in urban traffic spatio-temporal data,a three-layer urban traffic spatiotemporal data model with logical network layer,geometric network layer and traffic event characteristic layer is designed to store and manage traffic network facilities data and traffic system operating state data in an integrated way.The road network and dynamic flow data visualization methods are studied.In visualization of a road network,through parallel numeric method based on vector grid technology to classify symbol and classification symbol design is used to describe the way of technology level and the attribute values,dynamic traffic flow at the same time in the process of visualization,through a balance of visual content of dynamic scaling technique to realize the multi-scale map expression.Experimental results show that this method has higher visualization efficiency and more reasonable visualization effect than the general symbol library method,and avoids the symbol structure being destroyed in the process of map scaling.In the visualization of dynamic traffic flow data,the following three aspects are studied.Aiming at the visualization problem of traffic OD(Origin-destination)flow data,based on the basic idea of edge binding,a visualization method of traffic OD flow based on the optimal path network as constraint skeleton is proposed,which solves the problems such as the readability of existing visualized graph decreases significantly with the increase of data volume and poor visual effect.Aiming at the problem of visualization of traffic state at intersections,a thermal map generation method based on Bresenham’s circle drawing method and twodimensional Gaussian function is proposed to avoid the influence of factor density on analysis results in common density analysis methods.In addition,the paper introduces the spatial relationship characteristics of intersections and associated roads,and uses road network distance to replace the traditional Euclidean distance and Manhattan distance to analyze intersection hot spots.Compared with the common hot spot analysis algorithm,the obtained hot spot analysis graph is more consistent with the characteristics of traffic network.Finally,the spatio-temporal data management and visualization software of urban traffic is designed and developed,which realizes the functions of data management,data query and data visualization,and verifies the feasibility of the research method.